General AI Engineer interview prep
2 live roles at PleoLLMs, prompt engineering, RAG and AI application development — the fast-growing generative-AI engineering interview at Indian product startups, GCCs and AI-first companies. Distinct from a classic ML engineer: the focus is building on top of foundation models, not training them.
Typical AI Engineer interview rounds
- 1 LLM & GenAI fundamentals. How transformers and LLMs work, tokens, embeddings, context windows and prompting.
- 2 RAG & application design. Retrieval-augmented generation, vector databases, chunking and grounding.
- 3 Coding round. Python coding plus building an API around an LLM with a framework like LangChain.
- 4 System design & evaluation. Designing a production AI feature, handling cost, latency, safety and evaluation.
Commonly-asked AI Engineer questions
- What is the difference between fine-tuning, prompt engineering and retrieval-augmented generation?
- Explain how RAG works end to end, from a user query to a grounded answer.
- What are embeddings and how are they used in semantic search?
- What is a vector database and how does similarity search work?
- What is a context window, and how do you handle inputs that exceed it?
- What causes hallucinations in LLMs and how do you reduce them?
- What is the difference between temperature and top-p sampling?
- How would you chunk documents for a RAG pipeline, and why does chunk size matter?